Baseband delay cumulative summation beam forming method, system and device for ultrasonic imaging and medium

Abstract

The invention relates to a baseband delay cumulative summation beam forming method, system and device for ultrasonic imaging and a medium, and the method comprises the steps: carrying out the collection and channel signal orthogonal demodulation of ultrasonic data, and carrying out the preprocessing, and obtaining a channel signal after the delay alignment phase rotation of each emission angle of each voxel in an imaging visual field; baseband delay multiplication summation operation is carried out on the channel signal after delay alignment phase rotation of each emission angle of each voxel, and a baseband DMAS output signal of an imaging view is obtained; and outputting a signal based on the baseband DMAS of the imaging visual field, and carrying out ultrasonic imaging. The method can be widely applied to the technical field of signal processing, and is particularly suitable for two-dimensional or three-dimensional ultrasonic positioning microscopic imaging, micro blood flow imaging or other imaging.

Description

Technical Field

[0001] This invention relates to the field of signal processing technology, and in particular to a baseband delay-multiply-sum (DMAS) beamforming method, system, device and medium for ultrasound imaging. Background Technology

[0002] Abnormalities in the structure and function of brain microvessels are closely related to various cerebrovascular diseases and neurodegenerative diseases. Accurate assessment of microvascular morphology and blood flow is of great significance for disease diagnosis, monitoring, and mechanistic research. Existing imaging methods have varying degrees of limitations: CT angiography (CTA) / magnetic resonance angiography (MRA) can cover the entire brain but are costly, may involve ionizing radiation, and have resolutions mostly at the millimeter level; optical coherence tomography (OCT), two-photon imaging, and other optical methods have high resolution but limited penetration depth; traditional transcranial Doppler ultrasound techniques are sensitive to large blood vessels but cannot present the resolution and sensitivity required for microvascular networks.

[0003] Ultrasound-guided microscopy (ULM) treats microbubble contrast agents as isolated scatterers within blood vessels, combining frame-by-frame localization and cross-frame tracking to achieve micrometer-level super-resolution imaging at centimeter-level penetration depths. It has already shown application potential in organs such as the brain, heart, and kidneys. However, current ULM applications are mostly two-dimensional. Two-dimensional ULM suffers from limitations such as missing out-of-plane vascular and velocity components, projection errors, and operator dependence, making the development of three-dimensional ULM inevitable.

[0004] 3D ULM requires 3D channel data acquisition and volume reconstruction. Compared to the layer-by-layer acquisition method of mechanical scanning one-dimensional arrays, fully addressed two-dimensional matrix arrays can cover the entire volume at a higher volume frame rate. However, matrix arrays are often limited by factors such as reduced effective aperture and element sensitivity, and relatively low center frequency, which leads to deterioration of the point spread function (PSF) and a decrease in signal-to-noise ratio, thereby reducing the performance of microbubble detection, localization and tracking.

[0005] After acquiring channel data, a typical ULM workflow includes beamforming, clutter filtering, and microbubble localization and tracking. Beamforming methods significantly impact microbubble localization and separation capabilities: a wide main lobe makes it difficult to separate adjacent microbubbles, while strong side lobes can lead to side lobes being misidentified as microbubbles. Traditional Delayed Summation (DAS) suffers from a wide main lobe and high side lobes; while coherence weighting methods such as coherence factor (CF) and spatial angular coherence factor (SACF) can improve PSF performance, they may suppress or discard effective microbubble signals in low signal-to-noise ratio or weak microbubble echo scenarios, thus affecting subsequent localization and tracking.

[0006] Delayed Multiplicative Summation (DMAS) is a nonlinear beamforming method that enhances spatial coherence by multiplicatively coupling delayed channel signal pairs, thereby suppressing sidelobes and enhancing target echoes. However, existing DMAS implementations are mostly in the radio frequency domain, often requiring oversampling and additional filtering to recover the spectral components after nonlinear multiplication, and are subject to limitations in... Each element needs to be approximately The computational complexity of this pairwise operation is O(n log n). It is difficult to directly use it for real-time or near-real-time processing of three-dimensional ULM in large-channel fully addressable arrays. Summary of the Invention

[0007] The purpose of this invention is to provide a method, system, device, and medium for baseband delay-multiplication summation beamforming for ultrasound imaging, in order to at least solve one of the following technical problems:

[0008] (1) To address the problem that the signal-to-noise ratio and sensitivity of fully addressable two-dimensional matrix arrays are relatively low, making it difficult to detect weak echoes of microbubbles, a beamforming method that can enhance the detectability of microbubbles is provided. (2) Traditional RF domain DMAS requires oversampling and additional bandpass filtering, and its computational complexity is... To address this problem, a method is provided to implement DMAS in the complex baseband domain with an equivalent simplified computational form, thereby reducing computational complexity.

[0009] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a baseband delay-multiplication summation beamforming method for ultrasound imaging, comprising: The ultrasound data is acquired and the channel signals are orthogonally demodulated, and the channel signals after delay alignment and phase rotation of each voxel's emission angle in the imaging field are obtained after preprocessing. For each voxel, the delayed, aligned, and phase-rotated channel signal is subjected to baseband delay multiplication and summation to obtain the baseband DMAS output signal of the imaging field of view. Ultrasonic imaging is performed based on the baseband DMAS output signal of the imaging field of view.

[0010] Furthermore, the acquisition of ultrasound data and orthogonal demodulation of channel signals, followed by preprocessing to obtain the delayed, aligned, and phase-rotated channel signals for each voxel's emission angle in the imaging field of view, includes: Based on the determined ultrasonic beam emission mode, an all-addressable array transducer is used to acquire ultrasonic data and obtain the radio frequency channel signals received by each array element. The radio frequency channel signals received by each array element are quadrature demodulated to obtain complex baseband channel signals; Clutter filtering is applied to the complex baseband channel signal to suppress tissue echoes and preserve microbubble scattering signals; For each voxel in the imaging field of view, interpolation and phase rotation are performed on the filtered complex baseband channel signal based on the calculated total propagation delay to obtain the delayed-aligned and phase-rotated channel signal of each voxel at each emission angle in the imaging field of view.

[0011] Furthermore, based on the determined ultrasonic beam emission mode, ultrasonic data acquisition is performed using a fully addressable array transducer to obtain the radio frequency channel signals received by each array element, including: Determine the array configuration of the fully addressable array transducer, wherein the array configuration can be any one or any combination of linear array, convex array, phased array or fully addressable two-dimensional matrix array; Define the coordinate system and symbols for the fully addressable array transducer; Based on the determined ultrasonic beam emission mode, the fully addressable array transducer is controlled to emit ultrasonic beams, and the radio frequency channel signals received by each array element are obtained.

[0012] Furthermore, the orthogonal demodulation of the RF channel signals received by each array element to obtain the complex baseband channel signal is performed using the following formula:

[0013] In the formula, and Representing time frames The lower emission dimension is The array element dimension is The channel signal and the complex baseband signal; This is a low-pass filter operator used to preserve the baseband envelope; For demodulating the carrier frequency.

[0014] Furthermore, for each voxel in the imaging field of view, interpolation and phase rotation are performed on the complex baseband channel signal based on the calculated total propagation delay to obtain the time-delay-aligned, phase-rotated channel signal for each voxel at each emission angle in the complete imaging field of view, including: For each voxel in the imaging field of view, calculate the first... The ultrasonic wave at the first emission angle and the first Total propagation delay of array elements The calculation formula is:

[0015] In the formula, Voxel representation No. The emission delay of ultrasonic waves at a given emission angle; Voxel representation No. Array element receiving delay; The calculated total propagation delay is aligned using a preset interpolation operator. The calculation formula is as follows:

[0016] In the formula, This represents the channel signal after delay alignment following linear interpolation; Indicates based on total propagation delay Operators for linear interpolation; This represents the filtered complex baseband channel signal; Based on the delayed-aligned channel signal, carrier phase is compensated to obtain the first voxel of each... The ultrasonic wave at the first emission angle and the first Channel signals after delay alignment and phase rotation of array elements The calculation formula is:

[0017] In the formula, For demodulating the carrier frequency.

[0018] Furthermore, the baseband delay-accumulated summation operation is performed on the channel signals after delay alignment and phase rotation for each voxel's emission angle, to obtain the baseband DMAS output signal of the imaging field of view, including: For each voxel The ultrasonic wave at the first emission angle and the first The amplitude of the channel signal after phase rotation is corrected by aligning the delay of the array elements; Based on the conjugate multiplication of the amplitude-corrected channel signal pairs and summing by taking the real part, the baseband DMAS pairwise summation formula is obtained; Using the identity, the pairwise summation in the baseband DMAS pairwise summation formula is equivalent to the calculation of the cumulative quantity, thus obtaining the first value of each voxel. The ultrasonic wave at the first emission angle and the first The baseband DMAS output signal of the array element; For each voxel The baseband DMAS output signals at each emission angle are coherently combined to obtain the baseband DMAS output signal for each voxel. Based on the baseband DMAS output signal of each voxel, the baseband DMAS output signal of the imaging field of view is obtained.

[0019] Furthermore, the ultrasound imaging based on the baseband DMAS output signal of the imaging field of view includes ULM imaging or micro-blood flow imaging based on the baseband DMAS output signal of the imaging field of view.

[0020] In a second aspect, the present invention provides a baseband delay-multiplication-summing beamforming system for ultrasound imaging, comprising: The data preprocessing module is used to acquire ultrasound data and orthogonally demodulate channel signals, and preprocess the data to obtain the delayed, phase-rotated channel signal of each voxel in the complete imaging field of view. The beamforming module is used to perform baseband delay multiplication and summation on the channel signal after delay alignment and phase rotation for each voxel, and obtain the baseband DMAS output signal of the complete imaging field of view. The image generation module is used for ultrasound imaging based on the baseband DMAS output signal with a complete imaging field of view.

[0021] Thirdly, the present invention provides a computer-readable storage medium for storing one or more programs, said one or more programs including instructions that, when executed by a computing device, cause the computing device to perform any method.

[0022] Fourthly, the present invention provides a computing device comprising: one or more processors and a memory, wherein the memory stores one or more programs and is configured to be executed by the one or more processors, the one or more programs including instructions for performing any method.

[0023] The present invention has the following advantages due to the adoption of the above technical solutions: (1) Enhanced microbubble detectability and number of tracking trajectories: In this invention, the baseband DMAS utilizes the spatial coherence of apertures to enhance microbubble signals, thereby increasing the number of effectively located and trackable microbubbles, thus improving the number of trajectories and the integrity of the vascular network, specifically as follows: Figure 5 (a) ~ Figure 5 As shown in (d); in in vivo rat brain experiments, approximately 2.34 million and 3.07 million trajectories were obtained at low and high microbubble injection rates, respectively, significantly higher than the comparative beamforming method, as detailed below. Figure 3 (a) ~ Figure 3 (h) such as Figure 4 (a) ~ Figure 4 As shown in (l); (2) Improved vascular saturation and convergence speed: Within the same acquisition duration, baseband DMAS can achieve higher vascular saturation, indicating that it can depict a larger proportion of microvascular networks in a shorter time; examples show that the saturation at the end of acquisition can reach about twice or more than that of the comparative method, specifically as follows: Figure 6 (a) ~ Figure 6 As shown in (b); (3) Improved ULM Effective Resolution: Although CF / SACF is superior in point target PSF index, DMAS achieves the highest or better results in effective spatial resolution based on Fourier shell correlation (FSC) assessment in kidney simulation, ULM phantom, and in vivo experiments, indicating that it can better improve the final ULM reconstruction quality. Specifically, as shown in the figure below. Figure 8 As shown.

[0024] (4) Significantly reduce the computational burden of DMAS: By simplifying the computational form, the DMAS for each voxel is transformed from a pairwise multiplication summation to a single traversal accumulation calculation, thus reducing the complexity from Downgraded to In the embodiment, an acceleration of approximately 370 times is achieved on the GPU, demonstrating the engineering advantages of the method of the present invention.

[0025] (5) Avoid the common oversampling and additional filtering requirements of DMAS in the radio frequency domain: The present invention performs DMAS operation directly in the complex baseband domain, which can reduce the oversampling and bandpass filtering overhead introduced by DMAS in the radio frequency domain to process the spectral components after multiplication, and further improve the feasibility of engineering. Therefore, the present invention can be widely applied in the field of signal processing technology, especially suitable for two-dimensional or three-dimensional ultrasound localization microscopy (ULM), micro-blood flow imaging or other imaging, and is particularly suitable for volume ULM imaging scenarios of large-channel fully addressable arrays (e.g., fully addressable two-dimensional matrix arrays). Attached Figure Description

[0026] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts. In the drawings: Figure 1 This is a flowchart of the baseband delay-accumulation summation beamforming method for ultrasound imaging provided in this embodiment of the invention; Figure 2 This is a schematic diagram of DMAS calculation provided in an embodiment of the present invention; Figure 3 (a) ~ Figure 3 (h) is the reconstructed ULM volume obtained from rat brain in vivo experiments using different beamforming methods at a relatively low MB injection rate (150 μL / min) as provided in this embodiment of the invention; wherein, Figure 3 (a) This is an example of using the DAS beamforming method. Figure 3 (b) To use the CF beamforming method, Figure 3 (c) The SACF beamforming method is used. Figure 3 (d) The DMAS beamforming method of this invention is used. Figure 3 (e) ~ Figure 3 (h) represents the four beamforming methods used respectively. Figure 3 (a) An enlarged view of the area within the white box; Figure 4 (a) ~ Figure 4(l) is the reconstructed ULM volume obtained from an in vivo experiment in rat brain at a relatively high MB injection rate (250 μL / min) provided in this embodiment of the invention; wherein, Figure 4 (a) This is an example of using the DAS beamforming method. Figure 4 (b) To use the CF beamforming method, Figure 4 (c) The SACF beamforming method is used. Figure 4 (d) The DMAS beamforming method of this invention is used. Figure 4 (e) shows the four beamforming methods. Figure 4 (a) An enlarged view of the area within the white box; Figure 4 (i) ~ Figure 4 (l) represent the four beamforming methods respectively. Figure 4 (a) An enlarged view of the area within the green box; Figure 5 (a) ~ Figure 5 (d) shows the trajectory length distribution and trajectory count in the in vivo rat brain experiment provided in this embodiment of the invention, wherein, Figure 5 (a) Trajectory length distribution in experiments with low MB injection rates (150 μL / min); Figure 5 (b) Trajectory length distribution in experiments with high MB injection rates (250 μL / min); Figure 5 (c) is the number of trajectories in experiments with low MB injection rates; Figure 5 (d) represents the number of trajectories in the high MB injection rate experiment; the y-axis in the figure uses a logarithmic scale to highlight the differences in long trajectories; Figure 6 (a) ~ Figure 6 (b) is a graph showing the relationship between the saturation curve of the ULM volume reconstructed using DAS, CF, SACF, and DMAS beamforming methods provided in this embodiment of the invention and the acquisition time. Figure 6 (a) represents a low MB injection rate; Figure 6 (b) represents a high MB injection rate; Figure 7 This is a sagittal view of the reconstructed ULM volume in rat brain in vivo experiment at a low MB injection rate (150 μL / min) provided in this embodiment of the invention; wherein, the first to fourth rows are the reconstruction results of DAS, CF, SACF and DMAS beamforming methods, respectively; the first to sixth columns are the reconstruction results of acquisition times of 100, 200, 300, 400, 500 and 600s, respectively. Figure 8 (a) ~ Figure 8 (b) is a schematic diagram of FSC analysis of the ULM volume reconstructed using DAS, CF, SACF, and DMAS beamforming methods provided in this embodiment of the invention, wherein... Figure 8 (a) represents a low MB injection rate; Figure 8 (b) represents a high MB injection rate. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention are within the scope of protection of the present invention.

[0028] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0029] In some embodiments of the present invention, a baseband delay-cumulative summation beamforming method for ultrasound imaging is provided. The method includes: transmitting ultrasound using a multi-angle plane wave or a virtual focus diverging wave; receiving echoes from all array elements and performing orthogonal demodulation to obtain a complex baseband signal; performing clutter filtering on the baseband channel signal to suppress tissue components and retain microbubble scattering signals; calculating the transmission and reception delays for each voxel, and obtaining the delayed-aligned and phase-rotated channel signal in the baseband domain through interpolation and phase rotation; performing baseband DMAS beamforming on this basis, and reducing the complexity per voxel by equivalently transforming the pairwise conjugate multiplication summation into a computational form based on cumulative quantities. Downgraded to Subsequently, microbubble localization, cross-frame tracking, and trajectory rendering are performed to output two-dimensional images or three-dimensional volume data. This invention significantly reduces the computational burden of DMAS while enhancing microbubble detectability, and can improve the trajectory density, vessel saturation, and effective spatial resolution of ULM, making it particularly suitable for three-dimensional ULM imaging with large-channel fully addressable arrays.

[0030] Correspondingly, in other embodiments of the present invention, a baseband delay-multiplication-summing beamforming system, apparatus, and medium for ultrasound imaging are provided.

[0031] Example 1 like Figure 1 , Figure 2As shown, this embodiment provides a baseband delay-multiplicative summation beamforming method for ultrasound imaging, which includes data acquisition, channel signal demodulation, clutter filtering, delay calculation and phase rotation, baseband DMAS beamforming, microbubble localization, microbubble tracking and ULM rendering, etc.

[0032] Specifically, it includes the following steps: 1) Acquire ultrasound data and perform orthogonal demodulation of channel signals, and preprocess to obtain the delayed, aligned, phase-rotated channel signals of each voxel in the imaging field of view at each emission angle; 2) Perform baseband delay multiplication and summation on the channel signal after delay alignment and phase rotation for each voxel at each emission angle, and obtain the baseband DMAS output signal of the imaging field of view; 3) Perform ultrasound imaging based on the baseband DMAS output signal of the imaging field of view.

[0033] Furthermore, step 1) above includes the following steps: 1.1) Based on the determined ultrasonic beam transmission mode, a fully addressable array transducer is used for ultrasonic data acquisition to obtain the radio frequency channel signals received by each array element. ; 1.2) Perform quadrature demodulation on the RF channel signals received by each array element to obtain the complex baseband channel signal. ; 1.3) For complex baseband channel signals Clutter filtering is performed to obtain To suppress tissue echoes and preserve microbubble scattering signals; 1.4) For each voxel in the imaging field of view, the complex baseband channel signal is evaluated based on the calculated total propagation delay. Perform interpolation and phase rotation to obtain the delayed, phase-rotated channel signal for each voxel at each emission angle. .

[0034] Furthermore, step 1.1 above includes the following steps: 1.1.1) Determine the array configuration of the fully addressable array transducer.

[0035] Specifically, a fully addressable array refers to an array structure in an ultrasound array where each element corresponds to a receiving channel. Each element corresponds to a receiving channel signal; this can mean that each element has an independent receiving channel, or that independent data acquisition for each element can be achieved through time-division / multiplexing, thus forming a channel data array that includes the element dimension.

[0036] In this embodiment, the array form of the fully addressable array transducer can be any one or any combination of linear array, convex array, phased array or fully addressable two-dimensional matrix array.

[0037] 1.1.2) Define the coordinate system and symbols for the fully addressable array transducer.

[0038] In this embodiment, the coordinate system of the fully addressable array transducer is: the array center plane is denoted as... The imaging depth direction is Positive direction of axis, the first The center position of each array element is denoted as The position of a voxel in the imaging field of view is denoted as .

[0039] 1.1.3) Based on the determined ultrasonic beam transmission mode, the fully addressable array transducer is controlled to transmit the beam, and the radio frequency channel signals received by each array element are obtained. .

[0040] In this embodiment, the fully addressable array transducer can emit various ultrasonic beams, including plane waves and / or divergent waves generated by a virtual focus, and may include multiple steering angles or multiple divergent directions. In implementation, it may include: a. Multi-angle plane wave emission, and coherent composite of beamforming results from different angles; and / or b. Emit divergent waves from multiple virtual focal points and coherently combine the beamforming results from different virtual focal points.

[0041] Furthermore, in step 1.2) above, the radio frequency channel signals received by each array element are orthogonally demodulated to obtain complex baseband signals:

[0042] In the formula, and Representing time frames The lower emission dimension (i.e., the emission angle, in this embodiment) is used as the lower emission dimension (i.e., the emission angle). (Taking one launch angle as an example for illustration) The array element dimension is The channel signal and the complex baseband signal; This is a low-pass filter operator used to preserve the baseband envelope; For demodulating the carrier frequency.

[0043] It should be noted that in this embodiment, orthogonal demodulation can employ IQ demodulation or other demodulation methods; this invention does not impose any limitations on this. All subsequent beamforming steps are implemented in the complex baseband domain to reduce computational complexity and avoid the oversampling and additional filtering operations required by DMAS in the RF domain.

[0044] Furthermore, in step 1.3) above, when performing clutter filtering on the complex baseband channel signal, it includes singular value decomposition filtering and / or high-pass filtering.

[0045] Specifically, the process of using singular value decomposition filtering is as follows: First, the complex baseband signal is divided into blocks according to time frames and a data matrix is ​​constructed, so that each data block contains several frames; Second, for each data block, tissue clutter is suppressed by discarding the components corresponding to low-order singular values; For example, in this embodiment, each data block contains 400 frames, and tissue clutter is suppressed by discarding the first 15 singular values.

[0046] High-pass filtering or a combination of high-pass filtering and singular value decomposition filtering can also be used to suppress low-frequency tissue motion components, resulting in the filtered product. .

[0047] Furthermore, step 1.4 above includes the following steps: 1.4.1) For each voxel in the imaging field of view, calculate the... The ultrasonic wave at the first emission angle and the first Total propagation delay of array elements.

[0048] Specifically, for each voxel in the imaging field of view , No. The ultrasonic wave at the first emission angle and the first The formula for calculating the total propagation delay of array elements is:

[0049] In the formula, Voxel representation No. The emission delay of ultrasonic waves at a given emission angle; Voxel representation No. Array element receiving delay.

[0050] When using plane wave transmission mode, the transmit delay and receive delay are expressed as follows:

[0051] In the formula, Represents the plane wave A unit vector of emission angles; Indicates the first Array element positions; To indicate the position of a voxel, denoted as ; The medium sound velocity can be represented by measuring the medium temperature or by calibration, depending on the application scenario (such as a water tank, phantom, or inside an animal). This medium sound velocity can then be used for data acquisition and beamforming to reduce delay errors.

[0052] When using the divergent wave / virtual focus transmission mode, the transmission delay and reception delay are expressed as follows:

[0053] In the formula, This is the virtual focus position.

[0054] 1.4.2) Use a preset interpolation operator to perform delay alignment on the total propagation delay calculated in step 1.4.1).

[0055] Due to total propagation delay Typically not an integer multiple of the sampling period, this invention employs an interpolation operator to align the total propagation delay. The interpolation operator can be any of linear interpolation, spline interpolation, or sinc interpolation; this invention uses a linear interpolation operator. For example, perform delay alignment on complex baseband signals:

[0056] In the formula, This represents the time-aligned signal after linear interpolation, i.e., for voxels. The first value used for DMAS is extracted after alignment based on the total propagation delay. The launch angle is the first Channel signals of array elements; Indicates based on total propagation delay Operators for linear interpolation.

[0057] 1.4.3) Based on the delayed-aligned channel signal, the carrier phase is compensated to obtain the first voxel of each... The ultrasonic wave at the first emission angle and the first Channel signals after delay alignment and phase rotation of array elements .

[0058] To ensure that "demodulation first, then delay" and "delay first, then demodulation" are equivalent in result, this embodiment compensates for the carrier phase to obtain a channel signal with delay alignment and phase rotation. , is represented as:

[0059] When the system demodulation definitions are different (e.g., different exponent signs or time references), the phase compensation term can take an equivalent form, but in essence, they all compensate for the carrier phase difference to achieve phase rotation.

[0060] 1.4.4) Repeat steps 1.4.1) to 1.4.3) to obtain the channel signal after delay alignment and phase rotation for each voxel at each emission angle.

[0061] Furthermore, step 2) above includes the following steps: 2.1) For each voxel, the first... Channel signal after delay alignment and phase rotation at each emission angle Amplitude correction is performed to obtain .

[0062] For a given launch With voxels Simplified notation To achieve amplitude processing in the baseband domain that is consistent with or similar to that of traditional RF DMAS, delay-aligned phase rotation signals under single transmission or single angle are processed. Amplitude correction yields:

[0063] 2.2) Based on the amplitude-corrected channel signal pairs, perform conjugate multiplication and sum the real parts to obtain the baseband DMAS pairwise summation formula (conjugate coupling).

[0064] Baseband DMAS performs conjugate multiplication on the channel signal pairs and sums the real parts:

[0065] In the formula, Indicates a given launch With voxels Data after DMAS beamforming; , express conjugate coupling, Indicates complex conjugation; Indicates the number of array elements. This coupling contributes more to the phase-coherent (coherent) components, thereby enhancing the target and suppressing incoherent noise / sidelobes.

[0066] 2.3) Using the identity, the pairwise summation in the baseband DMAS pairwise summation formula is equivalent to the calculation of the cumulative quantity, thus obtaining the first value of each voxel. Baseband DMAS output signal at each transmission angle ; Traditional pairwise computation requires approximately The complexity is O(n) for multiple multiplications and accumulations. Each voxel of 1024 array elements requires approximately 523,776 pairwise operations, posing a significant engineering burden. In this invention, an identity is used to equate pairwise summation to cumulative quantity calculation, allowing...

[0067] but

[0068] Therefore, we get:

[0069] In implementation, only need to Perform a linear traversal and accumulate simultaneously and This allows the DMAS output signal of the voxel to be obtained, reducing the complexity from Downgraded to For example, this embodiment achieves an efficient speedup of approximately 370 times compared to explicit pairwise computation DMAS.

[0070] 2.4) Repeat steps 2.1) to 2.3), for each voxel. The launch angle obtained Coherent composite is performed to obtain the baseband DMAS output signal with a complete imaging field of view, i.e., single frame or single volume output.

[0071] Furthermore, in step 3) above, ultrasound imaging is performed based on the baseband DMAS output signal of the imaging field of view, including ULM imaging or micro-blood flow imaging.

[0072] This embodiment uses ULM imaging as an example for explanation, and includes the following steps: 3.1) Microbubble localization: A microbubble image of a frame or volume is formed based on the baseband DMAS output signal, and the center point of the microbubble in the microbubble image is located.

[0073] 3.2) Microbubble tracking: The microbubble localization results in continuous time frames are correlated and tracked to obtain the microbubble trajectory.

[0074] 3.3) ULM rendering: Generate two-dimensional super-resolution images and / or three-dimensional super-resolution volume data based on microbubble trajectories.

[0075] Furthermore, in step 3.1) above, microbubble localization can be performed using any one or any combination of weighted centroid method, Gaussian fitting, correlation matching or deep learning localization.

[0076] This embodiment uses the weighted centroid method as an example. The implementation process is as follows: an intensity threshold is set for each frame / volume, and local maxima are found as candidate microbubbles. Then, the weighted centroid method is used to obtain the sub-voxel center position. To control false detections and computational burden under high-density conditions, an upper limit can be set on the number of microbubbles located per frame; for example, no more than 900 microbubbles can be located per frame. Different beamforming methods can set different thresholds to compensate for sidelobe level differences, while other parameters remain consistent to ensure fair comparison.

[0077] Furthermore, in step 3.2 above, microbubble tracking can employ any one or any combination of the Hungarian algorithm, Kalman filtering, or multiple hypothesis tracking.

[0078] In this embodiment, the Hungarian algorithm is used for inter-frame matching, and Kalman filtering can be combined for velocity estimation and trajectory smoothing. To eliminate unreliable trajectories, a maximum link distance, a minimum number of consecutive frames, and a minimum trajectory length threshold can be set. For example, setting a maximum link distance of 0.2 mm (corresponding to a maximum speed of approximately 100 mm / s) will discard trajectories that last less than 10 frames or have a length shorter than 0.05 mm.

[0079] Furthermore, in step 3.3) above, after tracking is completed, the trajectory points or lines are accumulated onto a high-resolution grid to form a microbubble density map, resulting in a two-dimensional super-resolution image or three-dimensional super-resolution volume data. The three-dimensional data can be further displayed using maximum intensity projection (MIP) or arbitrary cross-sections to observe the microvascular network. For example, the accumulated microbubble density map can be used to render the three-dimensional ULM volume, and coronal MIP can be used for comparison.

[0080] Example 2 To evaluate the ULM performance of the proposed beamforming method under different physiological conditions, two in vivo experiments were conducted in this embodiment. All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of Tsinghua University. Two 8-week-old male Sprague-Dawley (SD) rats were used in this embodiment. Each rat was anesthetized with isoflurane during craniotomy and ultrasound imaging. A 3D-printed head mount was fixed to the intact skull portion to eliminate head movement during volumetric acquisition.

[0081] For the first rat, a microbubble suspension diluted twice with physiological saline was intravenously injected at a constant infusion rate of 150 μL / min using a syringe pump. For the second rat, the same microbubble suspension was injected at a higher rate of 250 μL / min to evaluate the performance of the beamforming method at high microbubble concentrations. The effective acquisition times for the low and high microbubble concentration experiments were 600 s and 1200 s, respectively.

[0082] like Figure 3 (a) ~ Figure 3 As shown in (h), the ULM volume was reconstructed using DAS, CF, SACF, and DMAS beamforming methods at a relatively low MB injection rate (150 μL / min). This was done to facilitate a direct comparison of fine vascular structures. Figure 3 (a) highlights a region of interest (ROI) in a white box, and the corresponding zoomed-in view ( Figure 3 (e) ~ Figure 3(h) was used to clearly present local microvascular details from different beamforming methods. To ensure fairness in visual comparison, all reconstruction results used the same rendering parameters and displayed the coronal maximum intensity projection. All beamforming methods stably recovered the major vascular structures, and the overall ULM quality was good, indicating that the selected injection rate and imaging settings provided suitable MB detectability for this experiment. Under these conditions, the vascular networks generated by DAS, CF, and SACF were similar; in contrast, DMAS revealed more microvascular details (h). Figure 3 (a) marked with a white triangle), and with higher brightness under the same rendering parameters, it indicates that more MB positioning information and trajectory data are retained in its reconstruction results.

[0083] like Figure 4 (a) ~ Figure 4 As shown in (l), the ULM volume reconstructed using DAS, CF, SACF, and DMAS beamforming methods at a high injection rate (250 μL / min) is presented using coronal MIP, and the rendering strategy is consistent with that of the low injection rate experiment. To facilitate intuitive comparison of fine vascular structures, Figure 4 In (a), the white and green boxes respectively outline the two ROIs, and the corresponding magnified views are as follows: Figure 4 (e) ~ Figure 4 (h) (white box ROI) and Figure 4 (i) ~ Figure 4 (l) (green box ROI) is shown to highlight local microvascular details under different beamforming methods. Compared with low injection rate experiments ( Figure 3 (a) ~ Figure 3 Compared to (h), the overall reconstruction quality decreased at higher injection rates, and fewer vessels were restored in deep brain regions. Although higher injection rates increase the number of microvessels (MBs) in the blood, high concentrations exacerbate MB overlap, violating the MB separation assumption, leading to discarded or inaccurate localization results, and consequently reducing the number of effective usable localizations. Under these conditions, DAS, CF, and SACF reconstructed fewer microvascular structures and had reduced ability to depict deep regions; in contrast, DMAS still recovered more microvascular structures (h). Figure 4 (a) marked with a white triangle), and the reconstruction result is brighter, indicating that its MB detectability and trajectory preservation ability have been improved.

[0084] like Figure 5 (a) and Figure 5 As shown in (c), the trajectory length probability distribution and the counting of trajectories of different lengths are presented in the in vivo experiment at a low MB injection rate. The trajectory distribution is strongly skewed, with short trajectories dominating and long trajectories being extremely rare. The trajectory statistics of DAS and CF are highly similar, and... Figure 3 (a) ~ Figure 3The phenomenon observed in (h) where the visual quality of both methods is comparable is consistent; compared to DAS, SACF has a slightly reduced number of tracks, especially long tracks. Notably, DMAS has the largest number of tracks, with a significantly increased proportion of long tracks: DAS has approximately 1.38 million tracks, CF approximately 1.49 million, SACF approximately 1.25 million, while DMAS has approximately 2.34 million tracks, roughly twice that of other beamforming methods. This higher tracking density (especially the increase in long tracks) confirms... Figure 3 (a) ~ Figure 3 (h) DMAS can provide more detailed observations of vascular networks.

[0085] Consistently, such as Figure 6 As shown in (a), FSC analysis shows that DMAS achieves the highest effective spatial resolution (20.5 μm), while DAS achieves 21.37 μm, CF achieves 21.59 μm, and SACF achieves 22.81 μm.

[0086] like Figure 5 (b) and Figure 5 As shown in (d), the trajectory length probability distribution and trajectory count are observed in the high injection rate experiment. The results are consistent with the trend in the low injection rate experiment: short trajectories dominate, and long trajectories are fewer. The statistical results for DAS, CF, and SACF are similar, and SACF has fewer long trajectories than DAS. Compared with the low injection rate experiment, the long trajectories under high injection rate are generally shorter, which is consistent with the increased MB overlap and decreased tracking continuity caused by high MB density. Importantly, compared with the low injection rate experiment, DMAS has a more significant gain in trajectory count compared to DAS, CF, and SACF: the total number of trajectories for DAS is approximately 1.06 million, for CF approximately 1.08 million, for SACF approximately 910,000, while the total number of trajectories for DMAS is approximately 3.07 million, about three times that of other beamforming methods. This significant increase in trajectory density explains... Figure 4 (a) ~ Figure 4 (l) The effect of DMAS on improving vessel mapping. Although the acquisition time of the high MB concentration experiment was twice that of the low MB concentration experiment, the total number of DAS, CF and SACF trajectories was reduced, indicating that high MB concentration reduces the number of effective usable locations and reliable trajectories; in contrast, DMAS can still obtain more trajectories at high injection rates, but the increase is less than twice (about 1.5 times that of the low injection rate experiment), indicating that DMAS has stronger robustness to high MB concentration, but is still affected by the increase in MB density.

[0087] like Figure 6As shown in (b), FSC analysis shows that DMAS achieves the highest effective spatial resolution (18.87 μm), while DAS achieves 20.64 μm, CF achieves 20.47 μm, and SACF achieves 19.98 μm.

[0088] like Figure 7 As shown, the cumulative reconstruction results (sagittal plane) are obtained as the acquisition time increases from 100 s to 600 s at a low MB injection rate (150 μL / min). The first to fourth rows represent the ULM volumes reconstructed using DAS, CF, SACF, and DMAS beamforming methods, respectively; the first to sixth columns represent the reconstruction results at acquisition times of 100, 200, 300, 400, 500, and 600 s, respectively. For all beamforming methods, the vascular network gradually becomes more complete with increasing acquisition time, reflecting the improving effect of more MB localization and tracking data on microvascular depiction. At each acquisition time point, DMAS exhibits higher vascular saturation and richer microvascular detail.

[0089] like Figure 8 As shown in (a), the saturation curve further confirms this: the DMAS curve is consistently higher than other beamforming methods throughout the acquisition time, indicating faster convergence and higher overall saturation; at the final acquisition time point, the saturation of DMAS is approximately twice that of other beamforming methods. As the acquisition time increases, the growth rate of the saturation curve decreases because the proportion of newly explored vessels gradually decreases.

[0090] like Figure 8 As shown in (b), the saturation curves of the ULM volume reconstructed using four beamforming methods at high MB injection rates are related to the acquisition time, and the trends are similar to those shown in (b). Figure 8 (a) Consistency: DMAS achieves the highest saturation at all time points, and at the final acquisition time point, its saturation is about three times that of other beamforming methods; as the acquisition time increases, the growth rate of the saturation curve decreases slightly.

[0091] Overall, compared with other beamforming methods, the DMAS beamforming method provides superior ULM image quality (including spatial resolution and vessel saturation).

[0092] Example 3 The above-described embodiment 1 provides a baseband delay cumulative multiplication and summation beamforming method for ultrasound imaging. Correspondingly, this embodiment provides a baseband delay cumulative multiplication and summation beamforming system for ultrasound imaging. The system provided in this embodiment can implement the baseband delay cumulative multiplication and summation beamforming method for ultrasound imaging of embodiment 1. This system can be implemented through software, hardware, or a combination of both. For example, the system may include integrated or separate functional modules or units to perform the corresponding steps in the methods of embodiment 1. Since the system in this embodiment is basically similar to the method embodiment, the description process in this embodiment is relatively simple. Relevant details can be found in the description of embodiment 1. The system embodiment provided in this embodiment is merely illustrative.

[0093] This embodiment provides a baseband delay-multiplication-summing beamforming system for ultrasound imaging, comprising: The data preprocessing module is used to acquire ultrasound data and orthogonally demodulate channel signals, and preprocess the data to obtain the delayed, phase-rotated channel signal of each voxel in the complete imaging field of view. The beamforming module is used to perform baseband delay multiplication and summation on the channel signal after delay alignment and phase rotation for each voxel, and obtain the baseband DMAS output signal of the complete imaging field of view. The image generation module is used for ultrasound imaging based on the baseband DMAS output signal with a complete imaging field of view.

[0094] Example 4 This embodiment provides a processing device corresponding to the baseband delay multiplication summation beamforming method for ultrasound imaging provided in Embodiment 1. The processing device can be a client-side processing device, such as a mobile phone, laptop, tablet, desktop computer, server, etc., to execute the method of Embodiment 1.

[0095] The processing device includes a processor, a memory, a communication interface, and a bus. The processor, memory, and communication interface are connected via the bus to enable communication between them. The memory stores a computer program that can run on the processor. When the processor runs the computer program, it executes the baseband delay-multiplication summation beamforming method for ultrasound imaging provided in Embodiment 1.

[0096] Preferably, the memory may be high-speed random access memory (RAM), and may also include non-volatile memory, such as at least one disk storage device.

[0097] Preferably, the processor can be any type of general-purpose processor such as a central processing unit (CPU) or a digital signal processor (DSP), and there is no limitation herein.

[0098] Example 5 The baseband delay cumulative summation beamforming method for ultrasound imaging described in Embodiment 1 can be specifically implemented as a computer program product. The computer program product may include a computer-readable storage medium on which computer-readable program instructions for executing the baseband delay cumulative summation beamforming method for ultrasound imaging described in Embodiment 1 are loaded.

[0099] A computer-readable storage medium can be a tangible device that holds and stores instructions for use by an instruction execution device. A computer-readable storage medium can be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any combination thereof.

[0100] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0101] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0102] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0103] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0104] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A baseband delay-multiplication summation beamforming method for ultrasound imaging, characterized in that, include: The ultrasound data is acquired and the channel signals are orthogonally demodulated, and the channel signals after delay alignment and phase rotation of each voxel's emission angle in the imaging field are obtained after preprocessing. For each voxel, the delayed, aligned, and phase-rotated channel signal is subjected to baseband delay multiplication and summation to obtain the baseband DMAS output signal of the imaging field of view. Ultrasonic imaging is performed based on the baseband DMAS output signal of the imaging field of view.

2. The baseband delay-multiplication summation beamforming method for ultrasound imaging as described in claim 1, characterized in that, The process of acquiring ultrasound data and orthogonally demodulating the channel signals, and preprocessing to obtain the delayed, aligned, and phase-rotated channel signals for each voxel's emission angle in the imaging field of view, includes: Based on the determined ultrasonic beam emission mode, an all-addressable array transducer is used to acquire ultrasonic data and obtain the radio frequency channel signals received by each array element. The radio frequency channel signals received by each array element are quadrature demodulated to obtain complex baseband channel signals; Clutter filtering is applied to the complex baseband channel signal to suppress tissue echoes and preserve microbubble scattering signals; For each voxel in the imaging field of view, interpolation and phase rotation are performed on the filtered complex baseband channel signal based on the calculated total propagation delay to obtain the delayed-aligned and phase-rotated channel signal of each voxel at each emission angle in the imaging field of view.

3. The baseband delay-multiplication summation beamforming method for ultrasound imaging as described in claim 2, characterized in that, Based on a defined ultrasonic beam emission mode, ultrasonic data acquisition is performed using a fully addressable array transducer to obtain the radio frequency channel signals received by each array element, including: Determine the array configuration of the fully addressable array transducer, wherein the array configuration can be any one or any combination of linear array, convex array, phased array or fully addressable two-dimensional matrix array; Define the coordinate system and symbols for the fully addressable array transducer; Based on the determined ultrasonic beam emission mode, the fully addressable array transducer is controlled to emit ultrasonic beams, and the radio frequency channel signals received by each array element are obtained.

4. The baseband delay-multiplication summation beamforming method for ultrasound imaging as described in claim 2, characterized in that, The formula for obtaining the complex baseband channel signal by orthogonally demodulating the radio frequency channel signal received by each array element is as follows: In the formula, and Representing time frames The lower emission dimension is The array element dimension is The channel signal and the complex baseband signal; This is a low-pass filter operator used to preserve the baseband envelope; For demodulating the carrier frequency.

5. The baseband delay-multiplication summation beamforming method for ultrasound imaging as described in claim 2, characterized in that, For each voxel in the imaging field of view, interpolation and phase rotation are performed on the complex baseband channel signal based on the calculated total propagation delay to obtain the time-delay-aligned and phase-rotated channel signal for each voxel at each emission angle in the complete imaging field of view, including: For each voxel in the imaging field of view, calculate the first... The ultrasonic wave at the first emission angle and the first Total propagation delay of array elements The calculation formula is: In the formula, Voxel representation No. The emission delay of ultrasonic waves at a given emission angle; Voxel representation No. Array element receiving delay; The calculated total propagation delay is aligned using a preset interpolation operator. The calculation formula is as follows: In the formula, This represents the channel signal after delay alignment following linear interpolation; Indicates based on total propagation delay Operators for linear interpolation; This represents the filtered complex baseband channel signal; Based on the delayed-aligned channel signal, carrier phase is compensated to obtain the first voxel of each... Channel signal after delay alignment and phase rotation at each emission angle The calculation formula is: In the formula, For demodulating the carrier frequency.

6. The baseband delay-multiplication summation beamforming method for ultrasound imaging as described in claim 1, characterized in that, The baseband delay-accumulated summation operation is performed on the channel signals after delay alignment and phase rotation for each emission angle of each voxel to obtain the baseband DMAS output signal of the imaging field of view, including: For each voxel Amplitude correction is performed on the channel signal after phase rotation and time alignment at each emission angle; Based on the conjugate multiplication of the amplitude-corrected channel signal pairs and summing by taking the real part, the baseband DMAS pairwise summation formula is obtained; Using the identity, the pairwise summation in the baseband DMAS pairwise summation formula is equivalent to the calculation of the cumulative quantity, thus obtaining the first voxel of each voxel. Baseband DMAS output signal at each transmission angle; For each voxel The baseband DMAS output signals at each emission angle are coherently combined to obtain the baseband DMAS output signal for each voxel. Based on the baseband DMAS output signal of each voxel, the baseband DMAS output signal of the imaging field of view is obtained.

7. The baseband delay-multiplication summation beamforming method for ultrasound imaging as described in claim 1, characterized in that, The baseband DMAS output signal based on the imaging field of view is used for ultrasound imaging, including ULM imaging or micro-blood flow imaging based on the baseband DMAS output signal based on the imaging field of view.

8. A baseband delay-multiplication-summing beamforming system for ultrasonic imaging, characterized in that, include: The data preprocessing module is used to acquire ultrasound data and orthogonally demodulate channel signals, and preprocess the data to obtain the delayed, phase-rotated channel signal of each voxel in the complete imaging field of view. The beamforming module is used to perform baseband delay multiplication and summation on the channel signal after delay alignment and phase rotation for each voxel, and obtain the baseband DMAS output signal of the complete imaging field of view. The image generation module is used for ultrasound imaging based on the baseband DMAS output signal with a complete imaging field of view.

9. A computer-readable storage medium for storing one or more programs, characterized in that, The one or more programs include instructions that, when executed by a computing device, cause the computing device to perform any of the methods described in claims 1 to 7.

10. A computing device, characterized in that, include: One or more processors and a memory, wherein the memory stores one or more programs and is configured to be executed by the one or more processors, the one or more programs including instructions for performing any of the methods described in claims 1 to 7.

Images